ToiFall CSI-Movement Dataset

Syncope and strokes in toilets can lead to severe injuries, and even pose life threats to patients. ToiFall project collects Channel State Information (CSI) from commodity Wi-Fi devices. Different human movements form various textures on images drawn with CSI data, and such textures can be used for feature extraction and classification.

This dataset consists of 1750 pictures collected from 17 volunteers, whom are asked to perform one of the six movements that happen frequently in toilets. These movements are illustrated in Figure 1: walk, stand up, sit down, fall to the left, fall to the right and fall to the front. For more information, please refer to our poster on Ubicomp’ 18.


Figure 1: Six movements performed by volunteers.
Figure 1: Six movements performed by volunteers.


The ToiFall dataset is available at:

To reference ToiFall:

Ziqi Wang, Zhihao Gu, Junwei Yin, Zhe Chen and Yuedong Xu, “Poster: Syncope Detection in Toilet Environments using Wi-Fi Channel State Information”, submitted to ACM Ubicomp 2018.